The main goal of this research is to elucidate the synaptic organization of circuits in the ventral part of the mouse spinal cord in both health and disease. Using a mouse model of spinal muscular atrophy we have demonstrated for the first time that loss of synaptic inputs to motoneurons occurs very early in the disease process and precedes motoneuron cell death and neuromuscular abnormalities. These findings open the possibility of finding new therapies for the muscle weakness that characterizes this disease. We now have preliminary evidence that spinal neurons projecting to the cerebellum and receive monosynaptic input from motoneurons. If these cells are excitatory, they could mediate the activation of locomotor networks by ventral root stimulation. Therefore, we propose to determine their connections, transmitter phenotype and behavior during locomotor-like activity. We made the surprising observation that stimulation of the ventral roots can excite locomotor like activity and entrain disinhibited bursting in the neonatal mouse cord. This was surprising because no known circuitry existed to mediate these excitatory effects. These excitatory effects of motoneurons on spinal networks require both ionotropic and metabotropic glutamate transmission but not cholinergic transmission. This is consistent with our work showing that motoneurons release an excitatory amino acid at their central terminals in addition to acetylcholine. Finally, we will establish if stimulation of motor axons can activate glial cells within the motor nucleus and if such activation is involved in the excitation of spinal networks by ventral root stimulation. We have successfully used viruses to define the presynaptic inputs to hindlimb motoneurons and are now involved in a collaboration with Dr. Sylvia Arber to image the activity of defined last-order neurons to motoneurons during locomotor-like activity in the neonatal mouse spinal cord. Finally, we are continuing to use the nematode to study the synaptic architecture of as simple locomotor network we have developed a novel formulation that reveals a hitherto unsuspected segmentation in the locomotor network.